BJR|Open https://​doi.​org/​10.​1259/​bjro.

​20200019

Received: Revised: Accepted:

19 May 2020 15 June 2020 16 June 2020

Cite this article as:

Beer L, Jajodia A, Prosch H. Pearls and pitfalls in lung cancer staging. BJR Open 2020; 2: 20200019.

REVIEW ARTICLE

Pearls and pitfalls in lung cancer staging

1,2
LUCIAN BEER, 3ANKUSH JAJODIA and 1HELMUT PROSCH
1
Department of Biomedical Imaging and Image-­guided Therapy, Medical University of Vienna, Vienna, Austria
2
Department of Radiology and Cancer Research UK, Cambridge, England, UK
3
Rajiv Gandhi Cancer Institute and Research Centre, Delhi, India

Address correspondence to: Dr Lucian Beer

E-mail: ​lucian.​beer@​meduniwien.​ac.​at

ABSTRACT
Lung cancer is the third most common cancer in the UK and is the leading cause of death. Radiology plays a central
role in the diagnostic work-­up of patients with suspected and known lung cancer. Tumour assessment includes both
local staging, as well as distant staging. Local staging objectives include the assessment of technical resectability with
regard to the evaluation of tumour size and invasion of surrounding structures. Distant staging objectives aim to iden-
tify distant metastasis in lymphatic and extra lymphatic tissues. CT, positron emission tomography/CT, MRI, and ultra-
sound are routinely used imaging techniques for staging in patients with lung cancer. In this review, we will consider the
pitfalls of these examinations that radiologists potentially face during the work-­up of patients with lung cancer.

INTRODUCTION The eighth edition addresses several limitations of the

Radiologists are fundamentally involved in the diagnosis
and treatment decision-­ making in patients with lung
cancer and see patients with lung cancer at various stages of
their journey. The radiologist is frequently the first contact,
often at time of their initial diagnosis, which can either be
during lung cancer screening examinations or during diag-
nostic examinations. Patients are then discussed within the
multidisciplinary team (MDT) and in many cases, the next
step in the diagnostic process is tumour tissue sampling, in
which radiologists are also involved. Once the diagnosis of
lung cancer is established, patients are staged based on the
anatomical tumour extent. Finally, imaging examinations
are used for treatment response assessment and for evalu-
ation of disease recurrence. Each step comprises peculiar-
ities and challenges for radiologists. In this review, we will
focus on the pearls and pitfalls of lung cancer staging from
a radiology perspective.
Eighth edition of the TNM classification of
malignant tumours
Since 01 Jan 2017, the eighth edition of the TNM staging
system, as proposed by the International Association for
the Study of Lung Cancer (IASLC), should be used to stage
non-­small cell lung cancer (NSCLC) and small cell lung
cancer (SCLC). This staging system is based on a data-
base analysis of 94,708 cases donated from 35 sources in
16 countries around the world between 1999 and 2010.1 Of
these, 70,967 validated cases were used for final analysis.
seventh edition, although several issues persist.
In the TNM staging system, the T descriptor describes the
local tumour extent, the N descriptor the involvement of
hilar or mediastinal lymph nodes, and the M descriptor
intra- and extrathoracic distant metastases.
T-descriptors
The T descriptor indicates the size of the primary tumour
and its extension into neighbouring structures, such as
the chest wall or the mediastinum. Furthermore, the T
descriptor is also used to describe ipsilateral pulmonary
metastases.
Based on their size, tumours are categorized as T1 tumours
with a maximum diameter of less than 3 cm, T2 tumours
with a maximum diameter between 5 cm and 7 cm, and T3
tumours with a maximum diameter of more than 7 cm. T1
and T2 tumours are further subclassified in T1a, T1b, and
T1c, and T2a and T2b in 1 cm increments to allow an even
better prognostic classification.2
How to measure tumour size
The TNM manifest defines the way in which radiologists
should measure lung cancer for TNM staging. For solid
tumours, the single largest dimension measured in one of
the three standard planes (axial, coronal, sagittal) using
thin sections (1 mm) should be measured and used for
the T descriptor (Figure 1A).3,4 In part-­solid tumours, the

© 2020 The Authors. Published by the British Institute of Radiology. This is an open access article distributed under the terms of the Creative Commons
Attribution 4.0 International License, which permits unrestricted use, distribution and reproduction in any medium, provided the original author and source
are credited.
 BJR|Open
Figure 1. Axial non-­enhanced CT showing (a) a solid nodule in
the left lower lobe compatible with a biopsy-­proven invasive
adenocarcinoma. It is recommended that the solid part be
measured (black line). The solid part represents the invasive
component of the tumour and should be used to define the
T classification for staging. In (b), a part-­solid tumour in the
right upper lobe is shown. It is recommended that the solid
dimension (black line) of the tumour for the T-­descriptor be
used. In (c), a pure ground-­glass nodule is shown. The largest
dimension of the ground-­glass tumour should be measured
and should be classified as an adenocarcinoma in situ (cTis).
2 of 11 birpublications.org/bjro
Beer et al
largest tumour dimension, including the ground-­ glass (GG)
part, should be recorded, but the staging is based on the largest
diameter of the solid component. The solid component in part-­
solid, non-­mucinous, lung adenocarcinomas correlates with
an invasive adenocarcinoma pattern, while the GG component
correlates with a lepidic growth pattern.3
Pure GG lesions with the largest diameter between 0.5 cm and
3 cm are classified as T1is, indicating that such pure GG lesions
are most likely in situ tumours (Figure 1C). Pure GG lesions
smaller than 0.5 cm are not staged, as the probability of such
lesions being malignant is very low. Pure GG lesions larger than
3 cm are classified as T1a.
Tumour size should be recorded in centimetres and include
millimetre increments. It is recommended that thin slices (1 mm)
be used, as thick-­ slice reconstructions (1.5 to 5 mm) could
mask small solid components. These tumours would, therefore,
mistakenly be classified as pure GG tumours (Figure 2).
In the rare cases of part-­solid tumours with several solid compo-
nents, the radiologist should measure the long axis of the largest
solid component.3
Atelectasis
In tumours in which the obstructive effect of the tumour leads to
an atelectasis or a post-­obstructive pneumonia, a reliable deter-
mination of tumour size is not possible. Consequently, partial or
total atelectasis of a lobe or entire lung side or a post-­obstructive
pneumonia are defined as T2.
In patients with central tumours leading to an atelectasis, in
which a delineation of tumours is necessary to limit the radi-
ation field, PET/CT or PET/MRI are of help in identifying the
obstructing tumour5 (Figure 3).
Infiltration of the chest wall or the diaphragm
While an infiltration of the chest wall is defined as T3, an infil-
tration of the diaphragm is classified as T4 disease.2 Multiplanar
reconstructions (MPRs) using sagittal and coronal views should
always be used to assess an infiltration of these structures, as axial
reconstruction might understate diaphragmatic infiltration.6
Unequivocal signs of an infiltration include the evidence of
tumour masses that have infiltrated neighbouring structures.
In case of the chest wall, the best evidence are rib erosions or
a broad extension of tumour masses into the intercostal space.
Less reliable signs of an infiltration of neighbouring structures
are impaired respiratory movement or a thickening of the pleura.
Both of these signs, however, have only a limited positive predic-
tive value, as inflammatory reactions around tumours may also
result in the same signs.
MRI, especially using dynamic CINE sequences, has a higher
accuracy (77%) compared to CT (47%) for the evaluation of chest
wall and diaphragm infiltration.7 In particular, the sensitivity of
up to 100% for MRI is superior compared to the 60% sensitivity
BJR Open;2:20200019
 Review article: Pearls and pitfalls in lung cancer staging
Figure 2. Axial contrast-­enhanced CT shows an adenocarcinoma of the right upper lobe. Depending on the slice thickness, the
tumour displays pure ground-­glass properties on (a) 3 mm reconstructions or solid components on (b) 1 mm reconstructions. It is,
therefore, recommended to use only 1 mm reconstructions with sharp reconstruction kernels for lung cancer staging.
of conventional CT. Tumours that infiltrate the diaphragm and/
or chest wall do not follow the normal respiratory movement.7
This can be visualised on MRI and used for staging purposes.
Similarly, dynamic CT examinations have higher sensitivity and
specificity compared to conventional static CT.8
Superior sulcus tumours (Pancoast tumours) are classified as at
least T3 tumours because of their chest wall infiltration. In the
presence of invasion of the brachial plexus above C8, the verte-
bral body, spinal canal, or subclavian vessels, these tumours are
upstaged to T4.
Infiltration of the mediastinum
An infiltration of the mediastinum is defined as T4. The medi-
astinal structures that define T4 are the mediastinal fat, great
vessels, the oesophagus, the trachea, and the heart. While an
unequivocal invasion of the mediastinal fat is defined as T4,
an invasion of the mediastinal parietal pleura alone is defined
as T3.2 Therefore, tumour contact with the mediastinal pleura
without direct or indirect signs of invasion is not automatically
staged as T4. Tumour contact with a length of more than 3 cm
or an obtuse angle between the tumour and the mediastinum are
indirect signs of mediastinal infiltration. In contrast, direct infil-
tration of the mediastinal fat or of the structure contained within
the mediastinum (e.g. heart, oesophagus) are staged as T4.2 The
sensitivity and specificity for the assessment of mediastinal inva-
sion by CT ranges from 40 to 78% and 69 to 99% (summarised
in Seo et al,9 respectively), while it is up to 100 and 93% for cine
MR images.9
In the staging system, great vessels are defined as the aorta, the
superior and inferior vena cava, the main pulmonary trunk, and
the intrapericardial portions of the pulmonary arteries and veins.
While the aorta, the superior and inferior vena cava, as well as
the main pulmonary trunk can be identified easily on CT, the
border between the intra- and extrapericardial portions of the
pulmonary arteries and veins cannot, because the pericardial
3 of 11 birpublications.org/bjro
BJR|Open
fold cannot be visualised with CT. The pericardial fold around
the right pulmonary artery may be defined by an imaginary line
at the mid-­half of the superior vena cava, while the pericardial
fold of the left pulmonary artery is roughly 1 cm from the bifur-
cation of the main pulmonary artery (Figure 4).
Intrapulmonary metastases
Intrapulmonary metastases are classified based on their location.
Pulmonary metastases in the same lobe as the primary tumour
are classified as T3, while metastases in another lobe on the same
side as the primary as T4, and metastases to the contralateral
lung as M1a.2
While this classification of pulmonary metastases sounds
straightforward, it may lead to overstaging if the suspected
pulmonary metastases are not verified histologically, as most
pulmonary nodules detected in patients with lung cancer are
benign.10 Consequently, additional pulmonary nodules that
might have an impact on treatment should be verified histologi-
cally to determine whether they are malignant.
If malignant, the radiologist has to discriminate intrapulmonary
metastases from a second primary.
Table 1 summarises the radiological criteria by which to distin-
guish a second primary from a metastasis.10 Intrapulmonary
metastases should be considered for solid lung cancers that have
a separate tumour nodule(s) with a similar solid appearance and
with (likely) matching histologic appearance (Figure 5).
Second primary lung cancers should be staged separately, each
with a T, N, and M descriptor. For example, a patient with a
3.5 cm adenocarcinoma of the left upper lobe, a 2.4 cm squamous
cell carcinoma of the right lower lobe, and with a metastatic
lymph node in R11 without evidence of systemic metastasis,
should be classified as having a T2aN0M0 adenocarcinoma and
a T1cN1Mo squamous cell carcinoma.
BJR Open;2:20200019
 BJR|Open
Figure 3. (a) A 65-­ year-­old patient with NSCLC and right
upper lobe atelectasis. On the contrast-­enhanced CT (b) an
inhomogeneous mass is seen at the level of the right hilum
that obliterates the right upper lobe bronchus. The exact
tumour diameter cannot be assessed accurately. (c) FDG
-PET/CT shows focally increased FDG uptake in the tumour,
while the post-­stenotic atelectatic lung shows only a moderate
FDG uptake. In this patient, the metabolic information helps
to correctly assess tumour dimension. FDG, fludeoxyglucose;
NSCLC,non-­ small cell lung cancer; PET, positron emission
tomography
It is important to note that the above-­mentioned guidance for the
staging of primary lung cancer should not be applied in patients
with multiple subsolid tumour nodules (either pure GG nodules
or part-­solid nodules) or in patients with pneumonic-­type lung
cancer. For patients with multifocal GG or part-­solid nodules,
4 of 11 birpublications.org/bjro
Beer et al
Figure 4. (a) Axial contrast-­ enhanced CT of a 55-­ year-­ old
female patient with an NSCLC at the left hilum. As the distance
between the main pulmonary trunk and the tumour is more
than 10 mm, it is classified as T3 disease. (b) In contrast, in this
73-­year-­old male patient with NSCLC, the tumour is in direct
contact with the main pulmonary artery and left main pulmo-
nary artery, and therefore, staged as T4. NSCLC, non-­small
cell lung cancer.
the T descriptor is based on the highest T lesion (#/m) indicating
the number of tumour nodules. The T(#/m) classification should
be applied independently whether the GG/part-­solid lesions are
in the same or in different ipsi- or contralateral lobes. In contrast,
for pneumonic-­type lung adenocarcinoma, the T is based on size
for T3 if the tumour is limited to one single lobe, and T4 or M1a
Table 1. Criteria by which to distinguish second primary
tumours vs metastasis (adapted from Detterbeck t al10)
Relative criteria that favour synchronous tumours
Different radiography appearance (i.e. shape, density)
Different metabolic activity
Different growth rates (if previous imaging is available)
Absence of nodal or systemic metastasis
Relative criteria that favour metastasis
Same radiographic appearance
Similar growth rates (if previous imaging is available)
Significant nodal or systemic metastases
BJR Open;2:20200019
 Review article: Pearls and pitfalls in lung cancer staging BJR|Open

Figure 5. (a) Coronal and (b) axial contrast-­enhanced CT of
a female patient with three pure ground-­glass nodules that
represent in situ adenocarcinomas in the right and left upper
lobe, as well as the middle lobe (arrow). The correct staging
is based on the tumour dimension of the largest ground-­glass
nodules (3.3 cm), as well as the number of tumours, which is
given in brackets: T2a(3)N0M0.
Table 2 summarises potential pitfalls that can occur during
the T-­descriptor assessment and solutions to overcome these
limitations.
Nodal disease
The N-­ descriptor remained unchanged in the eighth TNM
edition compared to the seventh edition. Lymph node metastases
to the ipsilateral hilum are classified as N1, lymph node metas-
tases to the ipsilateral mediastinum or the subcarinal lymph
nodes as N2, and contralateral mediastinal or supraclavicular
lymph node metastases as N3.

Although CT is routinely used for the initial staging of patients

with lung cancer, it has limited accuracy for the detection of
thoracic lymph node metastases, as a short axis equal to or
larger than 1 cm of the lymph nodes is the only criterion used
to diagnose lymph node involvement. However, as also benign
lymph nodes, such as inflammatory lymph nodes, may exceed
this threshold, this criterion has only a low diagnostic value. In a
metanalysis, it could be shown that the pooled sensitivity of the
size criterion is 55%, with a sensitivity of 81%.11 By using this
criterion only, 42% of lymph nodes larger than 1 cm would be
overstaged, as they are benign, and 17% of metastases in lymph
nodes smaller than 1 cm would be missed.12

By using fludeoxyglucose positron emission tomography (FDG

if in different ipsilateral or contralateral lobes, respectively. Both
tumour types use a single N and M staging, regardless of how
many GG/part-­solid nodules are present.
PET)/CT, the sensitivity could be increased to 80%, with a spec-
ificity of 88%.11,13 Consequently, PET/CT is recommended for
patients planning to undergo curative therapy (surgery or radio-
therapy) to exclude lymph node involvement.14 However, as
FDG PET/CT also struggles with false-­positive results, histolog-
ical confirmation of enlarged and/or FDG-­positive lymph nodes
is required to confirm metastatic involvement.15 A negative
FDG PET in normal-­sized hilar or mediastinal lymph nodes in
patients with a small tumours (<3 cm) virtually excludes lymph
node metastases, and thus, further invasive staging is not neces-
sary. However, in larger (≥3 cm) central tumours, even a negative
FDG PET/CT does not exclude lymph node metastases and inva-
sive staging should be performed.15–17
Equally important to the diagnosis of lymph node involvement
is the correct assignment of the lymph node location. Based on

Table 2. Pitfalls for assessing the T stage

Pitfall Effect Solution

Thick slices for T-­descriptor assessment
(1.5–5 mm)
Axial reconstructions only
Using non-­lung window settings without
sharp filter
Assessing diaphragmatic infiltration on
axial reconstructions
MPR, multiplanar reconstruction.
The solid component in a part-­solid tumour could
be missed → misclassification as pure ground-­glass
tumours
Potentially not capturing the single largest tumour
dimension→ low T descriptor
Underestimation of tumour dimension → low T
descriptor
Diaphragmatic infiltration might be missed → low T
descriptor
Thin slices (1 mm)
MPR in axial, sagittal, and coronal
reconstructions to assess single largest tumour
dimension
Lung-­window setting with a sharp filter
Sagittal and coronal reconstructions for
assessment of diaphragmatic infiltration

5 of 11 birpublications.org/bjro BJR Open;2:20200019

 BJR|Open
Figure 6. (a) Axial and (b) coronal contrast enhanced CT showing and enlarged (>10 mm) lymph node in the station 10 on the right
sight (10R). Please note that the lymph node is below the level of the azygos vein.
the anatomic location, the involved lymph nodes are assigned
to lymph node stations that are defined in the TNM atlas.18
Common pitfalls in the assignment of the appropriate lymph
node station are the assignment of right vs left paratracheal
lymph nodes.19 Importantly, the border between the right and
left paratracheal lymph nodes is not the midline of the trachea,
but the left lateral wall of the trachea.18
A common misclassification of lymph nodes can also occur at the
border between the lower paratracheal lymph nodes and hilar
lymph nodes. On the right side, the lower aspect of the azygos
vein separates the lower paratracheal lymph nodes from the hilar
lymph nodes (Figure 6); on the left side, this border is defined
by the upper aspect of the main pulmonary artery.18 In case of
doubt, coronal reformations are helpful to identify the anatomic
landmarks to determine the appropriate lymph node station.
Some lymph node stations in the thorax (i.e. anterior, middle,
and posterior diaphragmatic nodes, intercostal nodes, internal
mammary nodes, retrocrural nodes, and axillary nodes) are not
included in the International Association for the Study of Lung
Cancer lymph node map at all. Therefore, it is unclear whether
they are classified as N3 or M1 disease (Figure 7). Axillary
lymph node metastases are seen in <1% of patients with NSCLC
at the time of presentation and are associated with other M1
features in ~50% of cases.20 In the authors’ personal experience,
however, lymph node metastasis in the above-­mentioned lymph
node stations rarely occurs without other evidence of advanced
tumour stage (i.e. tumour size, lymph node metastasis, distant
metastasis) that drive the tumour staging.
Metastatic disease
Approximately 20–50% of patients with lung cancer have distant
metastases at the time of initial staging, with bone, brain,
adrenal, and liver being the most common locations for metas-
tasis.21 Importantly, some studies have shown that metastasis can
6 of 11 birpublications.org/bjro
Beer et al
occur at the time of angiogenesis when lesions are as small as
1–2 mm.22,23
In the TNM staging system, distant metastases are described
using the M descriptor.
The M1a category describes intrathoracic metastases, such as one
or more additional nodules in the contralateral lung or a tumour
with malignant pleural or pericardial nodules or effusion.
Malignant pleural or pericardial effusion
Malignant pleural or pericardial effusion and or pleural or peri-
cardial metastases are defined as M1a. Up to 16% of patients
with NSCLC have malignant pleural effusion at presentation.24
The sensitivity and specificity of CT for reporting malig-
nant pleural effusion are 62–75%, and 72–78%, respectively.25
Imaging features that suggest malignancy are nodular pleural
thickening, mediastinal pleural thickening, and parietal pleural
thickening >1 cm. However, with a negative predictive value of
~65%, approximately one in every three patients with suspected
malignant pleural disease and pleural effusion, without any other
radiological signs of malignancy, will have underlying malignant
disease.25 Thus, in case of equivocal findings on CT, patients
should undergo pleural tap or invasive pleural biopsies.
Extrathoracic metastases
Extrathoracic metastases are subdivided into two subcategories,
namely, M1b and M1c. The M1b category describes one solitary
extrathoracic metastases in one single organ. M1c is defined
by the presence of two or more extrathoracic metastases. The
subclassification of extrathoracic metastases into M1b and M1c
was introduced in the eighth edition of the TNM staging system,
acknowledging the difference in survival rates between M1b and
M1c.26
BJR Open;2:20200019
 Review article: Pearls and pitfalls in lung cancer staging
Figure 7. Lymph node stations that are not specified in the
eighth TNM manifest include (a, b, c) internal mammary nodes,
(c) axillary nodes, and (d) diaphragmatic nodes (arrows).
They can either be classified as N3 or M1 disease.
7 of 11 birpublications.org/bjro
BJR|Open
Figure 8. Axial contrast-­ enhanced FDG-­ PET/CT of a
59-­year-­old female patient with NSCLC. The bone metastasis
in the right acetabulum was not visible on the (a) CT, while it
was clearly visible on the (b, c) FDG-­PET (arrow). The accu-
racy of the FDG-­PET/CT to detect bone metastasis is higher
compared to CT alone. FDG, fludeoxyglucose;PET, positron
emission tomography.
The reclassified descriptors not only provide an enhanced defi-
nition of metastatic disease and have a better ability to predict
prognosis, but also maintain the compatibility with the previous
existing descriptors of the seventh edition.
PET-­CT proved very informative about metastatic spread in
NSCLC, with an ability to detect unsuspected metastasis in up
to 28% of patients, and also had an impact on management in up
to 53% of patients.27
Bone metastases
Approximately 35% of patients with lung cancer will develop
bone metastasis during the course of their disease.28
BJR Open;2:20200019
 BJR|Open Beer et al

Figure 9. Coronal FDG PET/CT scan showing metabolically PET/CT may be seen in the case of osteoblastic formations, as
active, bilateral adrenal gland metastasis (arrow). FDG, flude- bone matrix proliferation reduces the glycolytic activity essential
oxyglucose;PET, positron emission tomography. for FDG uptake.

Adrenal gland metastases

Metastasis to the adrenal glands is common in patients with
NSCLC and is usually accompanied by metastases in other
organs, although it can present as oligometastatic disease.32 It is
essential to distinguish commonly occurring benign entities of
the adrenal gland, such as adenomas, from a metastatic lesion.
The prevalence of adrenal adenoma is reported to be related to
age; the frequency of unsuspected adenoma is 0.5% in patients
aged 20–29 years and 7% in those older than 70 years.33

In lipid-­rich adenomas, the non-­invasive diagnosis relies on

the proof of fat components. On CT, attenuation values below
10 Hounsfiedl unit (HU) in regions of interest (ROIs) encom-
passing two-­thirds of the circumference of the region are highly
specific for adenomas.33

As one-­third of adrenal adenomas have a low lipid content, CT

attenuation values above 10 HU do not exclude adenomas. To
diagnose lipid-­poor adenomas, contrast-­enhanced CT scans or

CT delivers a high spatial resolution of cortical and trabecular Figure 10. Axial a) in and b) opposed phase MRI images
bone to detect bone metastases. The availability of dedicated showing a right sided adrenal adenoma (signal drop in the
bone algorithms in acquisition protocols, the ability to adjust opposed phase comared to the in phase.
the window width and level, and the possibility of MPRs results
in a higher sensitivity for CT compared to plain radiography in
detecting both osteolytic and osteosclerotic metastases. Bone
metastasis in lung cancer can be osteolytic, osteoblastic, or
mixed, but are predominantly lytic on CT imaging. All morpho-
logic types of baseline metastatic lesions may become sclerotic
lesions if there is a therapeutic response. In contrast, if a ther-
apeutic response is not completely achieved, different patterns
are recognised. A meta-­analysis showed a higher sensitivity (92%
vs 87%) and specificity (98% vs 94%) for the detection of bone
metastases when CT was combined with 18F-­FDG PET.29

The accuracy of bone scintigraphy in the detection of bone

metastasis was 87 vs 98% for FDG-­PET/CT. PET scanning is
more sensitive and accurate than bone scanning for the detection
of skeletal metastases (Figure 8), and is, therefore, recommended
for all patients scheduled for curative surgery to exclude distant
metastases.14,30

Bone islands can mimic sclerotic metastasis on CT imaging, but

can usually be differentiated by their high attenuation values
measured on CT. Another commonly encountered pitfall are
haemangiomas of the vertebral body. The classical polka-­dot
appearance of haemangiomas may not be present ubiquitously.
Some haemangiomas may exhibit a tracer uptake on PET-­CT
and this may further add to confounding clinical questions. In a
majority of these instances when PET findings are positive and
CT is negative, MR imaging may be the problem-­solving tool.
Both T1 and more advanced techniques as a modified Dixon
TSE-­T2 sequences show excellent sensitivity and specificity for
the detection of bone metastasis.31 False-­negative results on FDG

8 of 11 birpublications.org/bjro BJR Open;2:20200019

 Review article: Pearls and pitfalls in lung cancer staging
Figure 11. Axial a) contrast-­enhanced CT and b) MRI of the neurocranium in a patient with newly diagnosed adenocarcinoma
who was neurologically asymptomatic at the time of imaging. The CT scan showed no abnormalities, whereas there was a focal,
increased contrast uptake in the left brainstem highly suggestive of a cerebral metastasis. The lesion was proven to be malignant
in the follow-­up examination by the pattern of growth.
MRI are performed. MRI using chemical shift imaging has a
higher sensitivity for the detection of lipids than CT, and is thus
indicated in adrenal lesions with attenuation values between
10 and 20 HU.33 Combined chemical-­shift and dynamic-­MR
imaging was shown to have a sensitivity of 91% and a speci-
ficity of 94% in the differentiation between benign and malig-
nant adrenal lesions.34 For adenomas with CT attenuation values
greater than 20 HU, washout CT is superior to chemical shift
imaging.33
PET/CT and MRI are useful in distinguishing benign and malig-
nant adrenal masses (Figures 9 and 10). In a metanalysis of nine
studies evaluating the accuracy of FDG PET/CT for the detection
of adrenal metastases in patients with NSCLC, the sensitivity was
89%, the specificity was 90%, the positive likelihood ratio was
8.5, and the negative likelihood ratio was 0.09.35 False-­negative
FDG PET/CT findings may be the result of haemorrhage or
necrosis in a metastasis measuring less than 1.0 cm.36
Brain metastases
Contrasted-­enhanced MRI has a higher sensitivity for the detec-
tion of brain metastasis compared to CT and PET. In the March
2019 revisions of the NICE guideline, contrast-­enhanced CT in
patients with Stage II NSCLC having treatment with curative
intent is recommended.37 In contrast, in patients with Stage III
NSCLC having treatment with curative intent contrast-­enhanced
MRI is recommended. In addition, in patients with neurological
symptoms, MRI of the brain should be performed (Figure 11) .38
9 of 11 birpublications.org/bjro
BJR|Open
Liver metastases
Isolated liver metastases have been reported in approximately
3–4% of NSCLC patients.39 Although the characteristic CT
appearance in a triple-­phase scan is a hypodense, hypoenhancing
lesion, a subset of liver metastases may rarely manifest as hype-
renhancing lesions, particularly when there is small cell neuro-
endocrine differentiation in the primary lung cancer. This may
sometimes be confounded with a commonly occurring benign
entity of the liver, such as a haemangioma. Although MRI has a
higher accuracy in the detection of liver metastases compared to
CT,40 it is not routinely recommended.38
Soft tissue metastases
The lung, followed by the kidney and the colon, is the most
common primary carcinoma site that leads to clinically
recognised soft-­tissue metastases.41 Relevant differentials of
soft tissue metastases include injection site granulomas, ather-
omas, and neurogenic tumours. On PET-­CT, false-­positive
results may be caused by foci of brown fat, but the typical
location at the neck, paravertebral, mediastinal, and axillary
regions may provide a clue to the diagnosis.
CONCLUSION
The TNM in its eighth edition provides a detailed frame-
work by which to describe the anatomical extent of disease in
patients with lung cancer. Despite several advantages compared
to the former version, some limitations remain. It is important
for radiologists to recognise these limitations.42 In those cases,
BJR Open;2:20200019
 BJR|Open
radiologists should be aware of the following general rule stated
in the manuscript. “If there is doubt concerning the correct T,
N, or M category to which a particular case should be allotted,
the lower (i.e. less advanced) category should be chosen”.43
REFERENCES
1. Goldstraw P, Chansky K, Crowley J,
Rami-­Porta R, Asamura H, Eberhardt
WEE, et al. The IASLC lung cancer staging
project: proposals for revision of the TNM
stage groupings in the forthcoming (eighth)
edition of the TNM classification for lung
cancer. J Thorac Oncol 2016; 11: 39–51. doi:
https://​doi.​org/​10.​1016/​j.​jtho.​2015.​09.​009
2. Rami-­Porta R, Bolejack V, Crowley J, Ball
D, Kim J, Lyons G, et al. The IASLC lung
cancer staging project: proposals for
the revisions of the T descriptors in the
forthcoming eighth edition of the TNM
classification for lung cancer. J Thorac
Oncol 2015; 10: 990–1003. doi: https://​doi.​
org/​10.​1097/​JTO.​0000000000000559
3. Travis WD, Asamura H, Bankier AA,
Beasley MB, Detterbeck F, Flieder DB,
et al. The IASLC lung cancer staging
project: proposals for coding T categories
for subsolid nodules and assessment of
tumor size in part-­solid tumors in the
forthcoming eighth edition of the TNM
classification of lung cancer. J Thorac Oncol
2016; 11: 1204–23. doi: https://​doi.​org/​10.​
1016/​j.​jtho.​2016.​03.​025
4. Bankier AA, MacMahon H, Goo JM,
Rubin GD, Schaefer-­Prokop CM, Naidich
DP. Recommendations for measuring
pulmonary nodules at CT: a statement from
the Fleischner Society. Radiology 2017;
285: 584–600. doi: https://​doi.​org/​10.​1148/​
radiol.​2017162894
5. Gerbaudo VH, Julius B. Anatomo-­metabolic
characteristics of atelectasis in F-18 FDG-­
PET/CT imaging. Eur J Radiol 2007; 64:
401–5. doi: https://​doi.​org/​10.​1016/​j.​ejrad.​
2007.​08.​013
6. Beer L, Oberndorfer F, Silva M, Prosch H.
Radiological Signs of Tumor Dissemination.
The Thorax: Springer; 2020. pp. 35–46.
7. Kajiwara N, Akata S, Uchida O, Usuda J,
Ohira T, Kawate N, et al. Cine MRI enables
better therapeutic planning than CT in
cases of possible lung cancer chest wall
invasion. Lung Cancer 2010; 69: 203–8. doi:
https://​doi.​org/​10.​1016/​j.​lungcan.​2009.​10.​
016
8. Yamashiro T, Moriya H, Tsubakimoto M,
Nagatani Y, Kimoto T, Murayama S, et al.
Preoperative assessment of parietal pleural
10 of 11 birpublications.org/bjro
Applying this rule in case of uncertainty could give patient the
choice of doubt and enable the attempt of a curative treatment.
invasion/adhesion of subpleural lung
cancer: advantage of software-­assisted
analysis of 4-­dimensional dynamic-­
ventilation computed tomography. Eur
Radiol 2019; 29: 5247–52. doi: https://​doi.​
org/​10.​1007/​s00330-​019-​06131-w
9. Seo JS, Kim YJ, Choi BW, Choe KO.
Usefulness of magnetic resonance imaging
for evaluation of cardiovascular invasion:
evaluation of sliding motion between
thoracic mass and adjacent structures on
cine Mr images. J Magn Reson Imaging
2005; 22: 234–41. doi: https://​doi.​org/​10.​
1002/​jmri.​20378
10. Detterbeck FC, Nicholson AG, Franklin
WA, Marom EM, Travis WD, Girard
N, et al. The IASLC lung cancer staging
project: summary of proposals for revisions
of the classification of lung cancers with
multiple pulmonary sites of involvement in
the forthcoming eighth edition of the TNM
classification. J Thorac Oncol 2016; 11:
639–50. doi: https://​doi.​org/​10.​1016/​j.​jtho.​
2016.​01.​024
11. Silvestri GA, Gonzalez AV, Jantz MA,
Margolis ML, Gould MK, Tanoue LT, et al.
Methods for staging non-­small cell lung
cancer: diagnosis and management of lung
cancer, 3rd ED: American College of chest
physicians evidence-­based clinical practice
guidelines. Chest 2013; 143(5 Suppl):
e211S–50. doi: https://​doi.​org/​10.​1378/​
chest.​12-​2355
12. Alongi F, Ragusa P, Montemaggi P, Bona
CM. Combining independent studies of
diagnostic fluorodeoxyglucose positron-­
emission tomography and computed
tomography in mediastinal lymph node
staging for non-­small cell lung cancer.
Tumori 2006; 92: 327–33. doi: https://​doi.​
org/​10.​1177/​030089160609200412
13. Schmidt-­Hansen M, Baldwin DR, Hasler
E, Zamora J, Abraira V, Roqué I Figuls
M, Roque IFM. Pet-­Ct for assessing
mediastinal lymph node involvement in
patients with suspected resectable non-­
small cell lung cancer. Cochrane Database
Syst Rev 2014;: CD009519Cd009519..
doi: https://​doi.​org/​10.​1002/​14651858.​
CD009519.​pub2
Beer et al
14. Postmus PE, Kerr KM, Oudkerk M, Senan
S, Waller DA, Vansteenkiste J, et al. Early
and locally advanced non-­small-­cell lung
cancer (NSCLC): ESMO clinical practice
guidelines for diagnosis, treatment and
follow-­up. Annals of Oncology 2017; 28:
iv1–21. doi: https://​doi.​org/​10.​1093/​
annonc/​mdx222
15. Novello S, Barlesi F, Califano R, Cufer
T, Ekman S, Levra MG, et al. Metastatic
non-­small-­cell lung cancer: ESMO clinical
practice guidelines for diagnosis, treatment
and follow-­up. Ann Oncol 2016; 27(suppl
5): v1–27. doi: https://​doi.​org/​10.​1093/​
annonc/​mdw326
16. Reck M, Rabe KF. Precision diagnosis
and treatment for advanced non-­small-­
cell lung cancer. N Engl J Med 2017; 377:
849–61. doi: https://​doi.​org/​10.​1056/​
NEJMra1703413
17. Beer L, Hochmair M, Prosch H. Pitfalls
in the radiological response assessment of
immunotherapy. Memo 2018; 11: 138–43.
doi: https://​doi.​org/​10.​1007/​s12254-​018-​
0389-x
18. Rusch VW, Asamura H, Watanabe H,
Giroux DJ, Rami-­Porta R, Goldstraw
P, et al. The IASLC lung cancer staging
project: a proposal for a new international
lymph node map in the forthcoming
seventh edition of the TNM classification
for lung cancer. J Thorac Oncol 2009; 4:
568–77. doi: https://​doi.​org/​10.​1097/​JTO.​
0b013e3181a0d82e
19. El-­Sherief AH, Lau CT, Obuchowski
NA, Mehta AC, Rice TW, Blackstone
EH. Cross-­disciplinary analysis of lymph
node classification in lung cancer on CT
scanning. Chest 2017; 151: 776–85. doi:
https://​doi.​org/​10.​1016/​j.​chest.​2016.​09.​016
20. Satoh H, Ishikawa H, Kagohashi K,
Kurishima K, Sekizawa K. Axillary lymph
node metastasis in lung cancer. Med Oncol
2009; 26: 147–50. doi: https://​doi.​org/​10.​
1007/​s12032-​008-​9097-4
21. Eberhardt WEE, Mitchell A, Crowley
J, Kondo H, Kim YT, Turrisi A, et al.
The IASLC lung cancer staging project:
proposals for the revision of the M
descriptors in the forthcoming eighth
edition of the TNM classification of
BJR Open;2:20200019
 Review article: Pearls and pitfalls in lung cancer staging
lung cancer. J Thorac Oncol 2015; 10:
1515–22. doi: https://​doi.​org/​10.​1097/​JTO.​
0000000000000673 29.
22. Folkman J. Clinical applications
of research on angiogenesis. New
England Journal of Medicine 1995; 333:
1757–63. doi: https://​doi.​org/​10.​1056/​
NEJM199512283332608
23. Quint LE, Tummala S, Brisson LJ, Francis
IR, Krupnick AS, Kazerooni EA, et al. 30.
Distribution of distant metastases from
newly diagnosed non-­small cell lung
cancer. Ann Thorac Surg 1996; 62: 246–50.
doi: https://​doi.​org/​10.​1016/​0003-​4975(​96)​
00220-2 31.
24. Morgensztern D, Waqar S, Subramanian
J, Trinkaus K, Govindan R. Prognostic
impact of malignant pleural effusion at
presentation in patients with metastatic
non-­small-­cell lung cancer. J Thorac Oncol
2012; 7: 1485–9. doi: https://​doi.​org/​10.​
1097/​JTO.​0b013e318267223a
25. Hallifax RJ, Haris M, Corcoran JP, 32.
Leyakathalikhan S, Brown E, Srikantharaja
D, et al. Role of CT in assessing pleural
malignancy prior to thoracoscopy. Thorax
2015; 70: 192–3. doi: https://​doi.​org/​10.​
1136/​thoraxjnl-​2014-​206054
26. Tufman A, Kahnert K,
Kauffmann-­Guerrero D, Manapov F, 33.
Milger K, Müller-­Lisse U, et al. Clinical
relevance of the M1b and M1c descriptors
from the proposed TNM 8 classification
of lung cancer. Strahlenther Onkol 2017; 34.
193: 392–401. doi: https://​doi.​org/​10.​1007/​
s00066-​017-​1118-9
27. Gallamini A, Zwarthoed C, Borra A.
Positron emission tomography (PET) in
oncology. Cancers 2014; 6: 1821–89. doi:
https://​doi.​org/​10.​3390/​cancers6041821 35.
28. D'Antonio C, Passaro A, Gori B,
Del Signore E, Migliorino MR, Ricciardi
S, et al. Bone and brain metastasis in lung
cancer: recent advances in therapeutic
strategies. Ther Adv Med Oncol 2014;
11 of 11 birpublications.org/bjro
6: 101–14. doi: https://​doi.​org/​10.​1177/​
1758834014521110
Qu X, Huang X, Yan W, Wu L, Dai K. A
meta-­analysis of ¹⁸FDG-­PET-­CT, ¹⁸FDG-­
PET, MRI and bone scintigraphy for
diagnosis of bone metastases in patients
with lung cancer. Eur J Radiol 2012; 81:
1007–15. doi: https://​doi.​org/​10.​1016/​j.​
ejrad.​2011.​01.​126
Carr LL, Finigan JH, Kern JA. Evaluation
and treatment of patients with non-­small
cell lung cancer. Med Clin North Am 2011;
95: 1041–54. doi: https://​doi.​org/​10.​1016/​j.​
mcna.​2011.​08.​001
Hahn S, Lee YH, Suh J-­S. Detection
of vertebral metastases: a comparison
between the modified Dixon turbo spin
echo T2 weighted MRI and conventional
T1 weighted MRI: a preliminary study
in a tertiary centre. Br J Radiol 2018; 91:
20170782. doi: https://​doi.​org/​10.​1259/​bjr.​
20170782
Heinz-­Peer G, Hönigschnabl S, Schneider
B, Niederle B, Kaserer K, Lechner G.
Characterization of adrenal masses
using MR imaging with histopathologic
correlation. AJR Am J Roentgenol 1999; 173:
15–22. doi: https://​doi.​org/​10.​2214/​ajr.​173.​
1.​10397092
Schieda N, Siegelman ES. Update on CT
and MRI of adrenal nodules. AJR Am J
Roentgenol 2017; 208: 1206–17. doi: https://​
doi.​org/​10.​2214/​AJR.​16.​17758
Jana S, Zhang T, Milstein DM, Isasi
CR, Blaufox MD. Fdg-­Pet and CT
characterization of adrenal lesions in
cancer patients. Eur J Nucl Med Mol
Imaging 2006; 33: 29–35. doi: https://​doi.​
org/​10.​1007/​s00259-​005-​1915-8
Wu Q, Luo W, Zhao Y, Xu F, Zhou Q. The
utility of 18F-­FDG PET/CT for the diagnosis
of adrenal metastasis in lung cancer: a
PRISMA-­compliant meta-­analysis. Nucl Med
Commun 2017; 38: 1117–24. doi: https://​doi.​
org/​10.​1097/​MNM.​0000000000000757
BJR|Open
36. Yun M, Kim W, Alnafisi N, Lacorte L, Jang
S, Alavi A. 18F-­Fdg PET in characterizing
adrenal lesions detected on CT or MRI. J
Nucl Med 2001; 42: 1795–9.
37. Maconachie R, Mercer T, Navani N,
McVeigh G. Lung cancer: diagnosis and
management: summary of updated NICE
guidance. 364. BMJ: British Medical Journal
(Online); 2019.
38. Planchard D, Popat S, Kerr K, Novello S,
Smit EF, Faivre-­Finn C, et al. Metastatic
non-­small cell lung cancer: ESMO clinical
practice guidelines for diagnosis, treatment
and follow-­up. Annals of Oncology 2018;
29: iv192–237. doi: https://​doi.​org/​10.​1093/​
annonc/​mdy275
39. Kagohashi K, Satoh H, Ishikawa H,
Ohtsuka M, Sekizawa K. Liver metastasis at
the time of initial diagnosis of lung cancer.
Med Oncol 2003; 20: 25–8. doi: https://​doi.​
org/​10.​1385/​MO:​20:​1:​25
40. Scharitzer M, Ba-­Ssalamah A, Ringl H,
Kölblinger C, Grünberger T, Weber M,
et al. Preoperative evaluation of colorectal
liver metastases: comparison between
gadoxetic acid-­enhanced 3.0-­T MRI
and contrast-­enhanced MDCT with
histopathological correlation. Eur Radiol
2013; 23: 2187–96. doi: https://​doi.​org/​10.​
1007/​s00330-​013-​2824-z
41. Perisano C, Spinelli MS, Graci C,
Scaramuzzo L, Marzetti E, Barone C, et al.
Soft tissue metastases in lung cancer:
a review of the literature. Eur Rev Med
Pharmacol Sci 2012; 16: 1908–14.
42. Vlahos I. Dilemmas in lung cancer staging.
Radiol Clin North Am 2018; 56: 419–35.
doi: https://​doi.​org/​10.​1016/​j.​rcl.​2018.​01.​
010
43. Goldstraw P. International association for
the study of lung cancer staging manual in
thoracic oncology. Orange Park. Editorial Rx
Press 2009;.
BJR Open;2:20200019